Fault detection of wind turbine gearbox using thermal network modelling and SCADA data

B Corley,A Mcdonald,J Carroll

doi:10.1088/1742-6596/1618/2/022042

B Corley, A Mcdonald + Show 1 more

Open Access

https://doi.org/10.1088/1742-6596/1618/2/022042

Copy DOI

Abstract

This work uses a detailed understanding of the physics inside a wind turbine gearbox and SCADA temperature data as an alternative to data-driven techniques for fault detection. Thermal modelling based on the principles of heat transfer theory is used with the aim of understanding the thermal behaviour of a ‘healthy’ gearbox and use it to detect abnormal gearbox operating conditions. Data for turbines, ‘healthy’ and one month to fail, are analysed for two different failure modes to see if a fault can be detected in advance with the aim to improve physical understanding of wind turbine gearbox operation and condition monitoring techniques.

Highlights

The European Union has set a target for 2030 of at least 32% of energy from renewable sources [1] to meet its emissions reduction commitments under the Paris Agreement, to combat climate change
Internal and external environment and machine conditions can influence temperature measurements, so by analysing the evolution of temperature, it is important to note whether the increased temperature is from a fault, or from higher load
The temperature data has been binned into intervals of power and the mean and standard deviation of each bin is plotted

Summary

Introduction

The European Union has set a target for 2030 of at least 32% of energy from renewable sources [1] to meet its emissions reduction commitments under the Paris Agreement, to combat climate change. Operation and Maintenance (O&M) contributes up to 25% of the total levelised cost of wind energy, for onshore and up to 40% for offshore [2]. A predictive maintenance strategy optimises asset life and resources to minimise O&M costs. Condition monitoring refers to processes that enable early detection of faults, failures and wear of machinery before they reach a catastrophic or secondary-damage stage. It can extend asset life, enable better maintenance planning and logistics, and can reduce routine maintenance [4], with the intention of minimising downtime, and O&M costs while maximising production [5]

Methods

Results

Discussion

Conclusion